Math biology
• What is it?
– Use mathematics to answer biological
questions
– Use biology to pose interesting
mathematical questions
• “Mathematics is biology's next
microscope, only better; biology is
mathematics’ next physics, only
better” (Joel Cohen, PLoS Biology
2004)
Correlation of drug activity x gene
expression for 60 cancer cell lines
Scherf et al 2000
Outline
•
•
•
•
Sub-fields of math biology
Models for spatial patterning
What do math biologists actually do?
Resources
Sub-fields
• Almost every biological field
(bias toward human-health-related: systems biology)
• [computational/mathematical/quantitative] ×
{neurobiology, ecology, genetics, immunology …}
• Most mathematical fields
(bias towards applied math)
• Overlaps with statistics, computer science
(bioinformatics: genetics, comp sci, statistics, math)
Sub-fields
(from PubMed search)
Biology
• Evolution of domesticated
animals
• Molecular biology of seed
germination
• Neurobiology and brain energy
use
• Gene expression and retinal
angiogenesis
• Cancer metastasis
• Global infectious disease
dynamics
• …
Math (and statistics)
• Bayesian network analysis
• Differential equation models
• Neural network models
• Analysis of variance
• Logistic regression
• Markov chain Monte Carlo
• Agent-based models
… combinatorics, linear algebra,
ordinary and partial differential
equations, numerical analysis,
stochastic processes, game
theory, …
Spatial pattern formation
• Formation and
dynamics of spatial
patterns
• Within cells
• Among cells
• Among microbes
• In ecological
communities
Wikipedia
Turing patterns
• Alan Turing (1912-1954)
• Cryptographer, computer
scientist, mathematical
biologist (!)
• General theory of
morphogenesis
• What creates biological
pattern?
Independent
Turing’s model
Fuller (2010)
Turing’s model
• Short-range activation (small diffusion)
• Long-range inhibition (large diffusion)
2
¶X
¶
X
*
*
= a(X - X ) - b(Y -Y ) + DX 2
¶t
¶z
2
¶Y
¶
Y
*
*
= c(X - X ) - d(Y -Y ) + DY 2
¶t
¶z
From Turing (1952)
• “Figure 2 shows such a pattern, obtained in a few hours by a manual
computation …”
• “The outlines of the black patches are somewhat less irregular than
they should be due to an inadequacy in the computation procedure”
Animal coloration
EFBC Feline
Conservation
Center ("The
Cat House")
Animal patterns: theory
• Murray: as
wavelength
decreases
should move
from patches
 rings
 spots
Valais goat (http://www.goatworld.com)
From bodies to tails
• What if the
wavelength
stays the
same but
the domain
size
decreases?
A problem for the theory?
Malayan tapir: Denver Zoo, Edinburgh Zoo
TP in developmental biology
• increasing support from biologists for role in basic
developmental patterns (not just colour spots)
Sheth et al. 2012
Microbial ecology
• Unexplored world of biology
• Enormous biodiversity
• Complex communities
– Slime molds
– Biofilms (quorum sensing)
– Microbiome
• Partial differential equations, evolutionary
game theory
Dynamic patterns and cooperation
• Slime molds
(Dictostyleum discoideum)
• Alternate between
– single-celled individual
– multi-celled “slug” & “stalk”
• When food gets scarce,
send out cyclic AMP signals
• How do patterns arise?
(PDEs)
• What prevents cheating?
(Evolutionary game theory)
Slime mold model
• Too complicated to
show here
• Predicts which signaling
molecule drives
patterns
• Turning off that gene in
model destroys pattern
• … and in experiment!
• Palsson et al 1997
Left: wild-type, right: mutant
Microbial ecology: competition
• Some strains of E. coli produce colicin (antibacterial)
• Poisons neighbours
• Energetically expensive
Chao and Levin 1981
Colicin: spatial competition model
• Colicin is expensive
• Colicin-resistant but
non-producing strains
beat colicin producers
• Rock-paper-scissors
• Spatial structure and
discrete individuals
required for
coexistence
Durrett and Levin 1997
Resource use: game theory
• “Cooperators” produce
enzymes to break down
sugars and release
them into the
environment
• Model and experiment:
“cooperators” win at
boundary
Another puzzle
• Stock photo
(from aCBC programme
on antibiotic resistance )
• Is it real?
• What mechanisms are
required to produce
these patterns?
– Non-local dispersal?
– Non-local competition?
Ecology: tiger bush
• Vegetation pattern of
semi-arid regions
• Ranges from stripy to
patchy
• What’s going on?
• Too big/slow for
experiments
Tiger bush: conceptual model
Tiger bush: math model
• Another simple PDE model (Klausmeier 1999)
• Both diffusion (spread, ∂2u/∂x2 ) and advection
(directional movement, ∂w/∂x )
How does math biology actually work?
• Applied math; basic or applied biology
• Write down equations that describe how the
system works
• Try to solve them
• Fail
• Do something else
• Try to figure out what the results mean
• Estimate parameters
• Try to figure out what the results really mean
Techniques
• Approximation (e.g. Taylor
expansion; perturbation analysis)
• Analysis (prove existence and
uniqueness of solutions)
• Numerical solutions
• Stochastic models
• Search and optimization algorithms
• Statistical analysis;
parameter estimation
Equilibrium and stability analysis
• SIR
(susceptible/
infected/
recovered)
equations
S
I
R
dS
= -b SI
dt
dI
= b SI - g I
dt
dR
=gI
dt
Equilibrium analysis
•
•
•
•
Equilibrium (“trivial”, “disease-free”): I*=0
What if I is small but not zero?
dI/dt = (βS-γ)I
If I stays small, S approximately constant then
dI
ò I = ò (b S - g )dt
ln(I(t)) = (b S - g )t
I(t) = I(0)e( b S-g )t
I increases exponentially if βS>γ: equilibrium is unstable
Stability analysis: Turing equations
•
•
•
•
Write down equations
Homogeneous state is an equilibrium state
Allow small perturbation
Calculate growth or decay of perturbations
as a function of wavelength
• Find least stable wavelength: if it’s unstable,
then pattern grows (at least initially)
Courses
Essential
• Calculus (1XX3)
• Ordinary differential
equations (2C03)
• Linear algebra (1B03)
• Probability and statistics
(STATS 2D03, 2MB3)
• Intro to modeling
(3MB3)
• Some programming
course (1MP3 or ?)
Useful
• Numerical analysis (2T03)
• More calculus (1XX3)
• More ODEs (3F03)
• More lin alg (2R03)
• PDEs (3FF3)
• Stochastic processes (STATS
3U03)
• Analysis (3A03)
• Combinatorics (3U03)
• Math biology! (4MB3)
Things I’ve done
• Understanding measles
dynamics
• Eco-evolutionary
dynamics of virulence
• Spatial pattern
formation in ecological
systems, and feedback
on competition
• Spatial epidemic
patterns
• Using genetic markers
to estimate sea turtle
movements
• Movement behaviour of
Florida panthers
• Carbon cycling in
terrestrial ecosystems
MB resources @ Mac
• Biology & math honours programme
• Courses:
– MATH 1B03 (linear algebra), 2X03
(advanced calculus); 2R03 (differential
equations); 3MB3 (intro to modeling); BIO
3S03, 3SS3, 3FF3
• People:
– Math & stats: Bolker, Earn, Lovric,
Wolkowicz
– Bio: Bolker, Dushoff, Golding, Stone
– Biophysics: Higgs
– PNB: Becker
– probably others I’m forgetting!
• http://ms.mcmaster.ca/mathbiol/
More resources
• Recent journal articles:
tinyurl.com/biomath000
• K. Sigmund, Games of Life
• R. Dawkins, The Selfish Gene
• H. Kokko, Modeling for Field Biologists and
other Interesting People